Regional Road Development and Maintenance Project —Additional Financing (RRP MON 48186-007)

ECONOMIC ANALYSIS

A. Introduction

1. An economic analysis was undertaken for the rehabilitation of the –, Darkhan–Altanbulag, and Khuiten Valley– road sections. The result incorporates an update of the assessment carried out under the current project and an assessment of the additional financing project. The analysis considers the economic costs and benefits of the works using a standard evaluation approach, comparing the with- and without-project cases, with an appraisal period of 20 years from the first year of operations and an economic opportunity cost of capital of 9% in accordance with Asian Development Bank (ADB) guidelines.1 The base year for monetary values was updated to 2018.

B. Economic Background

2. The Ulaanbaatar–Darkhan–Altanbulag road forms part of the main north–south transit corridor across , linking the Russian Federation to the People’s Republic of China. Ulaanbaatar, with a population of 1.4 million in 2017, is the capital and main political, economic, and cultural center of the country. Darkhan is the third largest city in Mongolia, with a population of 84,000 in 2017. The city developed as an industrial center from 1961, with financial assistance from the former Soviet Union. It has a steel mill, building industry, textile industry, and educational establishments. Its importance as an industrial center is second only to Ulaanbaatar. It is also an important agricultural center. Sukhbaatar, toward the northern end of the corridor, is a trade and industrial center with a population of 23,000. Altanbulag is a settlement at the border with the Russian Federation and has a population of about 5,000. The corridor also provides access to the road to (Bayan–Undur), the second largest city of Mongolia with a population of 98,000. Erdenet is a new settlement established in the 1970s to exploit the large copper and molybdenum deposits. Arvaikheer lies on the main transport link to western Mongolia. It is a center for traditional crafts and animal husbandry, with a population of 32,000.

3. After a sharp slowdown in 2014–2016, driven by a fall in commodity prices and reduced foreign direct investment, the Mongolian economy recovered strongly in 2017 and 2018. Strong growth looks set to continue, although the economy remains vulnerable to commodity price shocks because of its reliance on the mining sector. Forecast growth of gross domestic product (GDP) to 2023 was obtained from the International Monetary Fund database (Table 1). Beyond 2023, constant growth of 4% per annum was assumed.

Table 1: Gross Domestic Product Growth Forecast, 2017–2042 (% per annum) 2018 2019 2020 2021 2022 2023 2024 … 2042 12.58 9.08 7.80 8.06 8.67 9.01 4.00 … 4.00 GDP = gross domestic product. Source: International Monetary Fund. 2018. World Economic Outlook (October 2018). Washington, DC.

C. Traffic

4. Traffic forecasts were based on an analysis of existing traffic and general traffic growth trends. Information on existing traffic by vehicle type was provided by the Ministry of Road and Transport Development for 2017 from seasonal counts at several locations along the project roads. The data were converted to annual average daily traffic (AADT) as shown in Table 2. While

1 ADB. 2017. Guidelines for the Economic Analysis of Projects. Manila. 2

2017 is the last year for which full data are available, traffic volumes showed strong growth trends in 2018.

Table 2: Observed Traffic Volumes (2017 AADT) Section Car SUV Minibus Bus Light Medium Heavy Total Truck Truck Truck Ulaanbaatar–Darkhan 3,444 861 206 113 484 184 154 5,446 Darkhan–Altanbulag 1,196 299 73 24 252 67 107 2,018 Khuiten V.–Arvaikheer 389 97 38 29 56 40 56 705 AADT = annual average daily traffic, SUV = sport utility vehicle. Source: Ministry of Road and Transport Development.

5. Traffic was forecast for 20 years beyond the first year of operations by applying factors based on forecast growth in GDP and observed elasticities of demand. An elasticity to GDP of 1.30 was assumed for passenger vehicles and 0.55 for freight vehicles, based on observed elasticities over the 10-year period (2006–2015), consistent with elasticities used in previous ADB road sector reports.2 No generated traffic was assumed since the existing road is not considered to deter travel and the works will not result in the provision of additional infrastructure.

6. Vehicle occupancies and loadings were observed in recent traffic surveys in Mongolia. The average occupancy of cars (including sport utility vehicles [SUVs]) is 3.0, minibuses carry on average 3.0 passengers (plus driver) and buses 9.9 (plus driver). Trucks are 45% fully loaded, 7% partially loaded, and 8% empty.

Table 3: Forecast Traffic Volumes (AADT) Section Car SUV Minibus Bus Light Medium Heavy Total /Year Truck Truck Truck Ulaanbaatar–Darkhan 2020 3,864 966 231 127 509 193 162 6,052 2030 6,535 1,634 391 214 638 243 203 9,858 2040 10,313 2,578 617 338 776 295 247 15,164 Darkhan–Altanbulag 2020 1,342 335 82 27 265 70 112 2,233 2030 2,269 567 139 46 332 88 141 3,582 2040 3,581 895 219 72 404 107 172 5,450 Khuiten Valley–Arvaikheer 2020 494 123 48 37 62 44 62 872 2030 905 226 88 67 81 58 81 1,506 2040 1,429 356 140 107 98 70 98 2,298 AADT = annual average daily traffic, SUV = sport utility vehicle. Source: Asian Development Bank estimates.

D. Project Costs

7. The project costs comprise the costs of physical works (widening, rehabilitation, asphalt concrete overlay, surface dressing, and climbing lanes). Other costs include costs of construction supervision and design costs and physical contingencies. The physical works are assumed to be spread across a 6-year implementation period (2020–2025) for each subproject. For the climbing- lane sections, a residual value of the works is added as a negative cost in the final appraisal year, based on an estimated economic life of 30 years. Financial costs are converted to economic costs by (i) excluding taxes, price contingencies, and financial charges; (ii) using a 0.80 shadow price

2 ADB. 2012. Mongolia Road Sector Development to 2016. Manila.

3

factor for low-skilled labor costs; and (iii) applying a shadow exchange rate factor of 1.028, calculated based on Mongolia’s trade balance using the domestic price numeraire. Additional maintenance costs are assumed to be incurred in the with-project case compared with the without-project case in proportion to the increased pavement width, based on an estimated annual cost of $3,000 per kilometer.

E. Project Benefits

8. The sources of economic benefits are savings in vehicle operating costs (VOCs), passenger value of travel time (VOT) savings, and accident cost savings. It is assumed that the first section of physical works will be completed and operational in 2023.

9. Vehicle operating cost savings. The VOC savings were calculated by applying unit VOCs to the total vehicle-kilometers on each section. VOCs vary according to the type of terrain, surface roughness, and type of vehicle. VOCs for different vehicle types and international roughness index (IRI) were estimated using the VOC module of the industry standard Highway Development and Management Model (HDM-4). The resulting VOCs for a range of IRIs are shown in Table 4.

Table 4: Vehicle Operating Costs ($/km, rolling terrain) Medium 4-Wheel Medium Light Medium Heavy IRI Car Drive Light Bus Bus Truck Truck Truck Paved road 2.0 0.25 0.66 0.24 0.32 0.24 0.43 0.71 8.0 0.29 0.87 0.28 0.39 0.28 0.52 0.86 12.0 0.33 1.09 0.33 0.46 0.34 0.61 1.00 18.0 0.40 1.34 0.43 0.57 0.44 0.76 1.25 IRI = international roughness index, km = kilometer. Source: Asian Development Bank estimates.

10. Value of time savings. Unit values of time were derived from the most recent average gross monthly wage rate of the country of MNT960,000,3 assuming that 23% of trips were made for work purposes with an average of 3.0 passengers per car and 9.9 passengers per bus.4 Following standard practice, trips made for work purposes were valued at the gross monthly wage rate plus employers' overheads (estimated at 50% of the gross wage), while other trips were valued at 30% of the net monthly wage rate. The resulting VOTs were calculated to be $3.71/vehicle-hour for cars and $12.25/vehicle-hour for buses. The unit VOTs were increased in line with forecast growth in GDP on an annual basis.

11. Accident cost savings. The value of benefits is based on the road safety assessment carried out in Mongolia in the current project by iRAP,5 which assumed a cost of $222,451 per fatality and $55,613 per serious injury. In 2017, 272 accidents occurred on the Ulaanbaatar– Darkhan road involving 16 fatalities and 108 serious injuries.8 The accident rate was assumed to reduce by 20% after improvement. Accident rates for the other sections were based on rates set

3 National Statistical Office of Mongolia. 4 As observed in the origin–destination survey carried out for the project preparatory technical assistance. 5 The iRAP road safety design methodology uses road-attribute risk factors to assess the likelihood that a crash will occur, and the severity of those that do occur, for both existing roads and road designs. The methodology involves an economic analysis of road safety countermeasure options, with benefits expressed in terms of the value of deaths and serious injuries prevented. 8 Not adjusted for underreporting, which is potentially significant.

4

out by DeJong, adjusted to the level of motorization in Mongolia. For these sections, accident rates for the improved road compared with the existing road were based on the assumption that a road with an IRI of 16 or more will have 40% more accidents than a road with an IRI of 2 or better (iRAP).9 Roads with an IRI of 2–16 are assumed to have a proportionate change in the accident rate. Additional benefits derive from the implementation of specific safety measures.

12. The assumptions made relating to overall travel speed, surface roughness, and accident rates are summarized in Table 5.

Table 5: Project Characteristics and Benefit Assumptions Section Scenario Road Speed Truck IRI Terrain Accident Rate/ Type (km/h) Speed Type Million Vehicles (% of light (km) vehicles) Ulaanbaatar– Without project Paved 70 80 6 Rolling 0.67 Darkhan With project Paved 77 80 3 Rolling 0.54 Darkhan– Without project Paved 70 80 4 Rolling 1.05 Altanbulag With project Paved 77 80 4 Rolling 1.01 Khuiten V.– Without project Paved 60 80 7 Flat 1.07 Arvaikheer With project Paved 77 80 3 Flat 0.99 IRI = international roughness index, km/h = kilometers per hour. Source: Asian Development Bank estimates.

F. Results of the Economic Cost–Benefit Analysis

13. The costs and benefits were assessed individually for Ulaanbaatar–Darkhan–Altanbulag and Khuiten Valley–Arvaikheer and for both roads combined to produce an overall economic internal rate of return (EIRR) and net present value (NPV). The results of the analysis are shown in Table 6. The Ulaanbaatar–Darkhan–Altanbulag road has an EIRR of 24.1% and may be considered economically viable. This is significantly lower that the EIRR found in the current project because of the higher costs but a relatively low value of additional benefits. The Khuiten Valley–Arvaikheer road has an EIRR of 12.8% and is economically viable as an independent project. The combined project has an EIRR of 22.6% and is considered viable overall. The benefits come 58% from VOC savings, 26% from time savings, and 16% from safety improvements. The economic indicators and streams of costs and benefits for the overall project are shown in Table 7. Table 6: Economic Indicatorsa Section EIRR NPV (%) ($ million) Ulaanbaatar–Darkhan–Altanbulag 24.1 217.23 Khuiten Valley–Arvaikheer 12.8 6.64 Combined 22.6 223.58 EIRR = economic internal rate of return, NPV = net present value. a NPV at 9% in 2018. Source: Asian Development Bank estimates.

9 iRAP. 2017. Analysis of Changes in the Rate of Severe Crashes for Typical Road Infrastructure Investments. Table 13.

5

Table 7: Economic Evaluation of Project Sections Combined ($ million) Year Costs Benefits Net Capital Maintenance Total VOC Time Accident Total Benefit NPV 2019 0.6 0.0 0.6 0.0 0.0 0.0 0.0 (0.6) (0.5) 2020 16.2 0.0 16.2 0.0 0.0 0.0 0.0 (16.2) (13.7) 2021 42.7 0.0 42.7 0.0 0.0 0.0 0.0 (42.7) (33.0) 2022 31.7 0.0 31.7 0.0 0.0 0.0 0.0 (31.7) (22.4) 2023 5.3 0.0 5.3 0.0 0.0 0.0 0.0 (5.3) (3.5) 2024 1.2 0.0 1.2 0.0 0.0 0.0 0.0 (1.2) (0.7) 2025 0.2 0.0 0.2 0.0 0.0 0.0 0.0 (0.2) (0.1) 2026 0.0 0.0 0.0 24.6 7.1 4.8 36.5 36.5 18.3 2027 0.0 0.0 0.0 25.9 7.8 5.2 38.9 38.9 17.9 2028 0.0 0.0 0.0 27.2 8.4 5.6 41.3 41.2 17.4 2029 0.0 0.0 0.0 28.4 9.1 6.1 43.6 43.6 16.9 2030 0.0 0.0 0.0 29.7 9.8 6.5 46.0 46.0 16.3 2031 0.0 0.0 0.0 31.3 11.1 7.4 49.7 49.7 16.2 2032 0.0 0.0 0.0 32.8 12.4 8.2 53.4 53.4 16.0 2033 0.0 0.0 0.0 34.4 13.7 9.0 57.1 57.1 15.7 2034 0.0 0.0 0.0 36.0 15.0 9.9 60.9 60.9 15.3 2035 0.0 0.0 0.0 37.5 16.3 10.7 64.6 64.6 14.9 2036 0.0 0.0 0.0 39.1 17.6 11.5 68.3 68.3 14.5 2037 0.0 0.0 0.0 40.7 18.9 12.4 72.0 72.0 14.0 2038 0.0 0.0 0.0 42.3 20.3 13.2 75.7 75.7 13.5 2039 0.0 0.0 0.0 43.8 21.6 14.0 79.5 79.4 13.0 2040 0.0 0.0 0.0 45.4 22.9 14.9 83.2 83.2 12.5 2041 0.0 0.0 0.0 47.0 24.2 14.9 86.1 86.0 11.9 2042 0.0 0.0 0.0 48.6 25.5 14.9 88.9 88.9 11.2 2043 0.0 0.0 0.0 50.1 26.8 14.9 91.8 91.8 10.6 2044 0.0 0.0 0.0 51.7 28.1 14.9 94.7 94.7 10.1 2045 0.0 0.0 0.0 53.3 29.4 14.9 97.6 97.6 9.5 2046 (30.0) 0.0 (30.0) 54.9 30.7 14.9 100.5 130.4 11.7 Net present value @ 9% 223.58 Economic internal rate of return 22.6% ( ) = negative, NPV = net present value, VOC = vehicle operating cost. Source: Asian Development Bank estimates.

14. The economic indicators were subjected to sensitivity analysis to test different scenarios of costs and benefits (Table 8). The EIRR remains well above 9% in all cases.

Table 8: Sensitivity Analysisa Scenarios EIRR NPV (%) ($ million) Base case 22.6 223.58 Investment cost 20% higher 20.5 208.79 Benefits 20% lower 20.0 164.63 Investment cost 20% higher, benefits 20% lower 18.1 149.81 EIRR = economic internal rate of return, NPV = net present value. a NPV at 9% in 2018. Source: Asian Development Bank estimates.